Re: [idn] An ignorant question about TC<-> SC

[Paul Hoffman / IMC <phoffman@imc.org>]

At 10:06 AM +0800 10/28/01, xiang deng wrote:
>Yes, multiple registrations in DNS is better than do nothing.

Good, we agree here.

>But, there are still some problems can not solve:

And here too!

>1). Just do one for the SC and one for the TC, it isn't enough. Because other
>person can register mixed TC/SC domain name.

If anyone on this list has said "just do one for each", they are 
incorrect. The right solution is "do as many as are needed". That 
number will change with names, of course.

>How to deal with the disputations of
>registration. Such disputations will not disappear, if we have no 
>real solution about it.
>it's policy issue.

Exactly right. It has nothing to do with the IETF. The IETF has never 
dealt with the exact same policy issue for *any* language, nor have 
we even started writing up policy proposals for any. Many examples 
have already been given, such as "color" and "colour", or "koln" and 
"köln". Although not identical, these types of examples have the same 
kinds of problems to end users as Traditional and Simplified Chinese 
do.

>2). multiple registrations in DNS can solve a part of the issue, How 
>to solve the
>delegation issue of subdomain and keep the consistency.
>it's technology issue.

Yes, but not one for the IDN WG. This affects all host delegations, 
not just ones that use internationalized names, and it has been a 
real-world problem for over a decade. If people have technical 
solutions for this very thorny issue, they should definitely bring 
them to the DNSEXT working group.

>3). if we provide multiple registration solution for customs, we 
>must guarantee the multiple
>records belong to one custom. but from technology view, we can not 
>guarantee it.

Exactly right. That is the way it is for all languages (including US 
English), and it will affect everyone the same when we 
internationalize the DNS. In fact, it will also affect everyone the 
same for every internationalized protocol, not just DNS. There is no 
protocol now that says "mail sent to thom@example.com should be 
delivered to tom@example.com" even though those two are more 
equivalent to English-speakers than traditional and simplified 
Chinese characters are to Chinese-speakers.

--Paul Hoffman, Director
--Internet Mail Consortium